Artificial Neural Networks Flow Behavior of Polymer in Thin Wall Cavity

碩士 === 大葉大學 === 機械工程研究所碩士班 === 94 === ABSTRACT Because of the development of science and technology now, the industry of the computer is different on day new night, make people's door notice that the computer derives out artificial intelligence (artificial intelligence ) and one kind of neur...

Full description

Bibliographic Details
Main Authors: Huang-MengWen, 黃孟文
Other Authors: Cheng-Hsien Wu
Format: Others
Language:zh-TW
Published: 2006
Online Access:http://ndltd.ncl.edu.tw/handle/22301576051011765719
Description
Summary:碩士 === 大葉大學 === 機械工程研究所碩士班 === 94 === ABSTRACT Because of the development of science and technology now, the industry of the computer is different on day new night, make people's door notice that the computer derives out artificial intelligence (artificial intelligence ) and one kind of neural networks (neural networks ) different from human brain, the using type neural network adopts and tries the way to imitate human neural work of brain, deal with the units and accept and stimulate and exchange information by a lot of, receive a certain result, made the proper response by the brain centre. One kind of neural networks can adopt the hardware or the software way to be set up , hardware way will look like the structure of the parallel processor , will let a lot of processors (processors ) link to each other together with the high-speed bus (bus ). The way of the software is to describe the node that linked to each other as a means of the material structure and algorithm of performing , linking among the nodes can have weight (weight ) , some nodes can accept the stimulus of introduction , some nodes can jointly output some meaningful results . So we imitate the transmission operation course of the neuron with the circuit diagram, then turn its transmission course into the training equation preface that we need and use VB (Visual Basic ) and Matlab and Moldflow simulation, in addition, the basic one flows strength , heat are spread, the material strength equation preface is changed into the expert system via VB, via real experiment with go bankrupt best parameter person who transmit neural network is it train in order to get trend of number value in order to finish the intact neural network system to go on. Keyword:neural networks, artificial intelligence, processors